I would like to see if I can sustain an argument that is is literally the definition of "technical debt" in it's purest form: The cost in inflation adjusted dollars to do this, probably exceeds the incremental upgrade cost had it been done across the lifetime of the system.
Does anyone have the chops to do the thought experiment? It means working on the labour and input costs for the time period, assuming a rollout and semi continuous upgrade through different ages of technology, as opposed to paying out a diamond bracelet worth of cost at the end to jump over technology intervening moments.
This is BS. I'm not an engineer but I'm sure one can create an emulation layer, connected to a SSD, put whatever data/software on it, and have this work. It certainly wouldn't cost $212M. I bet a college student could do this. Oh, look https://en.wikipedia.org/wiki/Floppy_disk_hardware_emulator, it's already been built. This smells of corruption.
They are upgrading the whole system, the floppy is just easiest-to-explain part.
> The new control system is supposed to be five generations ahead of what Muni is using now,
[...] challenges in maintaining expertise in 1990s programming languages have further encouraged the SFMTA to seek upgrades.
As the article points out, $212m will upgrade the entire train control system to a modern one (they admit to be "five generations" behind). Phasing out floppy disks is a consequence of this. They're cereinlynnot paying someone this much to just replace the floppy disk system but nothing else; that would be insane. The headline is a little flippant, and they don't mean it literally.
Many people on HN these days seem to reply without reading the article headline first...
The headline here is manipulative.
> The $212 million contract includes support services from Hitachi for "20 to 25 years,"
The $212m is a 20-year service contract for the new technology and the costs of installing it, not a one-time cost to remove the old floppy disks.
I would like to see if I can sustain an argument that is is literally the definition of "technical debt" in it's purest form: The cost in inflation adjusted dollars to do this, probably exceeds the incremental upgrade cost had it been done across the lifetime of the system.
Does anyone have the chops to do the thought experiment? It means working on the labour and input costs for the time period, assuming a rollout and semi continuous upgrade through different ages of technology, as opposed to paying out a diamond bracelet worth of cost at the end to jump over technology intervening moments.
In 1998 those disks were already outdated technology.
I guarantee you the dos setup will be sorely missed
This is BS. I'm not an engineer but I'm sure one can create an emulation layer, connected to a SSD, put whatever data/software on it, and have this work. It certainly wouldn't cost $212M. I bet a college student could do this. Oh, look https://en.wikipedia.org/wiki/Floppy_disk_hardware_emulator, it's already been built. This smells of corruption.
They are upgrading the whole system, the floppy is just easiest-to-explain part.
> The new control system is supposed to be five generations ahead of what Muni is using now, [...] challenges in maintaining expertise in 1990s programming languages have further encouraged the SFMTA to seek upgrades.
The goal and costs are to replace the entire control system not find a means of replacing a single legacy component within it.
As the article points out, $212m will upgrade the entire train control system to a modern one (they admit to be "five generations" behind). Phasing out floppy disks is a consequence of this. They're cereinlynnot paying someone this much to just replace the floppy disk system but nothing else; that would be insane. The headline is a little flippant, and they don't mean it literally.
Many people on HN these days seem to reply without reading the article headline first...
>Many people on HN these days seem to reply without reading the article headline first...
You sound like you're new to the internet.
Also it is $212 for 20-25 years of support. Not just some one time cost.
my lord.
[dead]